Numerical Optimization (Springer Series in Operations Research and Financial Engineering)

This is a must-read book for anyone who wishes to study Numerical Analysis on a professional level. This text will teach many fundamental principals of applied analysis by walking the reader through an extremely wide array of finite dimensional optimization algorithms.

This book is a well-written, outstanding reference for anyone interested in understanding, using, and/or implementing state-of-the-art techniques in nonlinear optimization. Ample attention is paid to both constrained and unconstrained problem types, with a healthy and refreshing emphasis on trust-region strategies, and modern SQP and Interior-Point algorithms. Sufficient detail is paid to most topics while overall perspectives are well-maintained. This book is the very best of its kind for its intended audience. I strongly recommend it.

The book is quite complete and goes directly to the point. if you ever need optimization in your design you will find it here. Simple and well presented. It has enough details about algorithmic performance and description that should be enough to implement. It is a book that you will never regret having it in your library. If you want something more theoretical use Nonlinear Programming by Bertsekas. If you want to use optimization in your programs use this.

The best text book on the various issues around steepest descent, conjugate gradient, Newtonian methods etc. Clearly show you why you still need to care about steepest-descent even though we were taught it is much slower than Newton or CG. Those that are practical oriented might have ignored the key role SD play in many methods to guarantee convergence (or progress). Very good write up on the Wolfe condition, Cauchy point, and trust region.

Not the easiest to read at times but a classic, solid math and algorithms book. Comprehensive appendix for the math used, detailed (enough) proofs for the material presented and a good balance between examples and theory.

This book covers all the frequently appearing optimization problems and the optimization methods. It provides you all the necessary details so that you can understand the mathematical backgrounds and also you can implement the algorithms by yourself. The book is balanced very well. Absolutely great. I love this book.

Trained a physicist before I went into applied math and computer science I thoroughly enjoy this book. It math is not light but not that heavy either. I wouldn't consider it math at graduate level as most of the material is applicable to an undergrad with 1 year of calculus and another of linear algebra. A bit more of statistical elaborations might be useful for a third edition, given the current focus on machine learning. Anyhow as a first book it is still excellent.

I am trying to get into research for machine leaning and statistics. This book is written so accessibly, though you really should be comfortable with the basics of calculus and linear algebra to be able to read it. You will get more out of the book if you sit in front of the computer coding the examples and exercises while reading it--after all, it is a numerical optimization book. While I'm sure there are more recent methods that aren't covered in the book, I believe the book gives you more than a sufficient foundation to learn those methods very quickly. The reason I am giving an unfavorable review is that the book's package was left out in the pouring rain all day while I was at work. It's a wonder it did not get wet. The delivery team should be ashamed of themselves for doing this, especially when there are dry spots to put the package. I wonder how many books are permanently damaged because of this negligence.